- Email: [email protected]
Concordance Between Two Enzyme Immunoassays for the Detection of Clostridium difficile Toxins
Archives of Medical Research 41 (2010) 92e96
ORIGINAL ARTICLE
Concordance Between Two Enzyme Immunoassays for the Detection of Clostridium difficile Toxins Jaime O. Herrera-Ca´ceres, Adria´n Camacho-Ortiz, Arturo Galindo-Fraga, Melissa Herna´ndez-Dura´n, Andrea Cordero-Rangel, Araceli Herna´ndez-Cruz, Miriam Bobadilla-del Valle, Alfredo Ponce-de-Leo´n, and Jose´ Sifuentes-Osornio Laboratory of Clinical Microbiology, National Institute of Medical Sciences and Nutrition Salvador Zubira´n, Mexico City, Mexico Received for publication October 21, 2009; accepted January 26, 2010 (ARCMED-D-09-00509).
Background and Aims. The diagnosis of Clostridium difficile-associated disease (CDAD) is based on the detection of toxins from stool samples. There are several immunoassays for this purpose. The aim of this study was to determine the concordance between the two immunoassays and their performance in comparison to the toxigenic culture as part of the initial evaluation of a suspected case of CDAD. Methods. All fecal samples submitted for detection of C. difficile toxins during a 5-month period to our laboratory were analyzed by two immunoassays, VIDAS Toxin CDA/B assay (BioMerieux) and ImmunoCard Toxins A/B (Meridian Bioscience). We cultured on cycloserine-cefoxitin-fructose agar and PCR was used for detection of toxigenic genes. Real-time PCR was performed directly from samples to detect the tcdC gene. Results. At the end of the study we processed 230 samples, 13 were positive using VIDAS CDA/B (5.6%), and 14 using ImmunoCard A/B (6.0%); k coefficient was 0.857. With ImmunoCard A/B we obtained a sensitivity of 80%, a specificity of 99%, positive predicitive value (PPV) 86% and negative predictive value (NPV) 98%, as compared to toxigenic culture. For VIDAS CDA/B we obtained a sensitivity of 90%, a specificity of 98%, PPV 69% and NPV 99%, compared to the same standard. There were seven undetermined results (3.0%) by VIDAS CDA/B. Five of these had a positive culture and all the patients had symptoms of CDAD. Considering these undetermined results as positive, we calculated a sensitivity of 93%, specificity of 97%, PPV 71% and NPV of 99% for this test, and a k of 0.856. Both immunoassays showed similar results and are suitable for the initial evaluation of patients with suspected CDAD. Conclusions. Our data suggest that an undetermined result of VIDAS CDA/B should be considered as positive if CDAD is suspected. Additionally, both immunoassays showed similar results and are suitable for the initial evaluation of patients with suspected CDAD. Ó 2010 IMSS. Published by Elsevier Inc. Key Words: Diarrhea, Toxins, Clostridium difficile, Nosocomial infections.
Introduction Approximately 3e29% of patients receiving antibiotics will develop diarrhea; moreover, Clostridium difficile will be the culprit agent in up to 10e25% of these cases (1). Address reprint requests to: Jose´ Sifuentes-Osornio, MD, Laboratorio de Microbiologı´a Clı´nica, Instituto Nacional de Ciencias Me´dicas y Nutricio´n Salvador Zubira´n, Vasco de Quiroga #15, C.P. 14000, Tlalpan, Me´xico D.F. Me´xico; Phone: (þ52) (55) 5487-0900, Ext. 2174; FAX: (þ52) (55) 5513-3945; E-mail: [email protected]
In recent years, complications and mortality associated with this bacteria have increased dramatically (2e4). To date, the cytotoxicity assay for detection of toxin B is considered the gold standard for diagnosis of C. difficile-associated disease (CDAD). However, this procedure is cumbersome, costly and the results are not available before 48 h. Several immunoassays have been developed in order to evaluate the presence of both A and B toxins in stool samples as a simpler and faster option (5). Two popular immunoassays are the VIDAS Clostridium difficile Toxin A/B assay
0188-4409/$esee front matter. Copyright Ó 2010 IMSS. Published by Elsevier Inc. doi: 10.1016/j.arcmed.2010.03.004
Immunoenzymatic Detection of Clostridium difficile toxins
(bioMerieux, Marcy l’Etoile, France) with good sensitivity (88.3%) and specificity (99.8%) values, a positive predictive value (PPV) of 98.1%, and a negative predictive value (NPV) of 98.4%, and the ImmunoCard Toxins A/B (Meridian Bioscience, Cincinnati, OH), which have shown a sensitivity of 55.4%, a specificity of 92.2%, a PPV of 44.4%, and a NPV of 94.9%, when compared with tissue culture cytotoxin B assay (CBA) (6,7). Stool culture for C. difficile is time consuming and requires anaerobic conditions as well as a special medium containing cycloserincefoxitin-fructose agar (CCFA) or the recently studied C. difficile-Brucella agar (8,9). The presence of C. difficile can be determined by other methods including multiplex polymerase chain reaction (M-PCR) to detect tcdA and tcdB genes from culture (toxigenic culture) (10) and realtime PCR (RT-PCR) directly from stool samples to detect the tcdC gene (regulatory gene) with good sensitivity and specificity (9). The aim of this study was to determine the concordance between the two immunoassays and their performance in comparison to the toxigenic culture as part of the initial evaluation of a suspected case of CDAD. Materials and Methods Setting and Time Period During a 5-month period (MaySeptember, 2008), all samples sent for detection of C. difficile toxins to the Laboratory of Clinical Microbiology of a 200-bed tertiary-care medical center located in Mexico City were analyzed by both immunoassays VIDAS CDA/B (bioMerieux) and ImmunoCard A/B (Meridian Bioscience) methods. Each immunoassay was performed following manufacturer recommendations. Results from VIDAS CDA/B were reported in relative fluorescence values (RFV). Culture Methods All stool samples were cultured for C. difficile using CCFA and blood sheep agar (BSA) medium in anaerobic conditions after alcohol shock (11). Upon arrival, we stored 2 g of feces at 80 C. The positive cultures were identified by Vitek ANI test cards (bioMerieux). Molecular Methods The isolates were tested for the presence of tcdA and tcdB genes using a modification of a PCR-method previously described (9). In short, the sequences of the primers used for amplification were tcdB-F, GAGCTGCTTCAATTGGAGAGA; tcdB-R, GTAACCTACTTTCATAACACCAG; tcdA-F, ATGATAAGGCAACTTCAGTGG, and tcdA-R, TAAGTTCCTCCTGCTCCATCAA. The cycling parameters were amplification of fragments of the tcdA gene (624 bp) and the tcdB gene (412 bp) by conventional PCR; the conditions of the multiplex PCR reaction were
93
final volume 25 ml, dNTPs 300 mM, MgCl2 3 mM, primers 25 pM and Taq polymerase 2.5 units; the cycling parameters were denaturation for 5 min at 94 C and amplification consisting of 1 min at 94 C, 1 min at 50 C, and 1 min at 72 C for 30 cycles. The presence of an amplified product was confirmed by agarose gel detection. Real-time PCR was used to confirm the presence of the tcdC gene from the frozen stool as previously described (9). This method has previously shown high sensitivity (86%) and specificity (97%). Briefly, this PCR assay detects the presence of the tcdC gene as well as the 18-bp, 39-bp, and other deletions found in the tcdC gene of C. difficile. We used the kit 296 (TIB MolBiol LLC, Adelphia, NJ) following the manufacturer’s recommendations. Positive and negative controls were included with each run. The positive control was C. difficile ATCC 9689 (wild type) and a clinical isolate C. difficile type 027 kindly provided by Dr. Ellen J. Baron with the 18-bp deletion. The negative control was a recent clinical isolate of C. sordelli. Ethics and Statistical Analysis This study was evaluated and approved by the Institutional Review Board. For concordance analysis we excluded samples with an insufficient amount of stool for proper analysis by both methods. We also excluded repeated samples from the same patient. Sensitivity, specificity, positive and negative predictive values were calculated. Coefficient of k was considered useful with a value $0.4. We used SPSS software v. 15. Results A total of 234 fecal samples were processed during the study period. Of these, four were discarded: two due to an insufficient amount of feces and two for being repeated samples. Two hundred and thirty samples were included in the final analysis. Using the ImmunoCard A/B 14 (6%) samples were determined as positive and with the VIDAS CDA/B 13 samples were positive (5.6%). Of the 230 cultures included in the study, only 13 had bacterial growth (5.6%). The presence of tcdA and/or tcdB genes were confirmed in the 13 isolates by M-PCR (one isolate was tcdA negative but tcdB positive). Real-time PCR showed the presence of the tcdC gene directly from 15 fecal samples, two of them with a negative culture; therefore M-PCR could not be done. With these results, we compared the ImmunoCard A/B and VIDAS CDA/B with the results from the anaerobic culture and the presence of tcdC gene. Using the latter as the standard method, we calculated sensitivity, specificity, PPV and NPV. The ImmunoCard A/B test showed 80, 99, 86, and 99%, respectively, and the VIDAS CDA/B test showed 90, 98, 69 and 99%, respectively. VIDAS CDA/B values were classified into three categories: positive, negative and undetermined as recommended by the manufacturer. Two hundred and ten samples
94
Herrera-Ca´ceres et al./ Archives of Medical Research 41 (2010) 92e96
were negative, seven were undetermined and 13 were positive. Then, the seven samples with undetermined results were repeated. Of these seven samples, two were undetermined by a second assay, one was positive, and four were negative upon retesting. A more detailed analysis of these seven patients found that all had diarrhea days before the diagnostic test (median 2 days, range 2e15 days) was processed, four had fever, one had an elevated peripheral white blood cell count, and all had a normal serum creatinine. All had also received antibiotics for at least 3 days during the 6 weeks prior to sample processing and were hospitalized for a minimum of 2 days, two were hospitalized at the intensive care unit, and five were receiving proton pump inhibitors. Additionally, we compared these seven samples with the results from the other diagnostic tools used, finding a positive culture in five samples. The presence of the toxinproducing genes was confirmed by the M-PCR assay for these five cultures (Table 1). Distribution of the initial RFV of the three groups according to the VIDAS CDA/B method is shown in Figure 1. The group of samples with a negative value showed a mean value and the one standard deviation values far below the negative cutoff value. Samples with an undetermined result had a trend towards higher values close to the positive cutoff value, and samples with a positive result showed values far above the positive cutoff value. With the initial results the k coefficient between ImmunoCard A/B test and VIDAS CDA/B test was 0.857, and after the inclusion of the seven undetermined VIDAS CDA/B results into the positive group, it was 0.856. When considering VIDAS CDA/B undetermined results as positive, we obtained a sensitivity of 93%, a specificity of 97%, PPV of 71% and NPV of 99%. In Table 2 we describe the performance of the three methods used in this trial. The numbers of all the possible combined results by VIDAS CDA/B, ImmunoCard A/B and toxigenic culture results are shown. We found 12 samples that were positive by all methods, four others had a positive VIDAS CDA/B result but both of the other methods were negative. Two samples had a positive result by both immunoassays but neither culture nor RT-PCR confirmed these results. Two more samples were positive by VIDAS CDA/B and by toxigenic culture but negative using ImmunoCard A/B. From all samples there was only one positive toxigenic culture that had a negative result by both immunoassays. Discussion CDAD has gained great importance during the present decade. Detection of toxins in stool samples is fundamental for the diagnosis (1). However, a significant number of hospitals in Mexico and in many other developing countries still lack a laboratory method for such a purpose. As previously reported, less than one third of the Latin American
Table 1. Description of the results of those 7 samples with an undetermined value using the VIDAS CDA/B test
Vidas CDA/B 1. Undetermined*/ Undetermined** 2. Undetermined*/ Undetermined** 3. Undetermined*/Positive** 4. Undetermined*/Negative** 5. Undetermined*/Negative** 6. Undetermined*/Negative** 7. Undetermined*/Negative**
Immunocard A/B
Toxigenic Culture
Negative
Positive
Positive
Positive
Positive Negative Positive Negative Negative
Positive Positive Positive Negative Negative
*First assay. **Re-test.
countries (7/25 countries recognized by the Pan-American Health Organization) have reported cases of CDAD (12). The results of this study show the performance and a high concordance (kappa 0.856) among two easy to perform immunoassays. Although we observed a moderate prevalence of CDAD (6.52%) during this study, both of these immunoassays were quite trustworthy when compared with a good standard using up-to-date molecular biotechnology (toxigenic culture). VIDAS CDA/B, has the advantage of quantitative results of fluorescence that are automatically interpreted as positive, negative or undetermined. Nevertheless, in order to use VIDAS CDA/B it is necessary to have the VIDAS automated system, that can be used for many other serologic tests and has proven to be trustful for these other tests (13e16). On the other hand ImmunoCard A/B is a methodologically simple that does not need any special equipment other than a standard 4e8 C refrigerator and a vortex mixer. For these reasons and the prompt results (up to 30 min), ImmunoCard A/B can be used by non-specialized personnel and needs no sophisticated laboratory environment. It has been also widely studied and has a proven credibility detecting both toxins, A and B (17,18). It has also been shown that well-type assays and membrane-type assays to detect these toxins do not differ significantly when compared with cytotoxicity as a reference standard. Welltype assays and membrane-type assays with glutamate dehydrogenase as a target perform equally well and have good correlation with feces culture of C. difficile. Furthermore in our study, the results by this method were corroborated by two observers. In three samples there was discordance among the observers and the test was repeated and given as negative by both on the second trial. Given the results from the sub analysis of the RFV obtained by the VIDAS CDA/B, it would be advisable to treat such a patient with specific therapy anti-CDAD if this patient has risk factors, develops diarrhea and has an undetermined result by the VIDAS CDA/B test. Using these
Relative Fluorescence Values
Immunoenzymatic Detection of Clostridium difficile toxins
95
6.22 6.07 6.01 4.15 3.94 1.95 1.53 1.41 1.39 1.37 0.53 0.51 0.49 0.34 0.32 0.24 0.21 0.2 0.16 0.13 0.11 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01
Negative
Undetermined
Positive
VIDAS CDA/B Result Figure 1. Distribution of relative florescence values from the VIDAS CDA/B Cutoff relative fluorescence values: Negative #14 units, Undetermined O14e37 units, Positive O37 units, as shown by the red doted lines. The mean value of the negative results was 0.02 (minimum 0.01- maximum 0.13, SD 0.022), the mean value of the undetermined results was 0.24 (0.16e0.34, SD 0.06), and the mean value of the positive results was 1.41 (0.49e6.22, SD 2.23).
criteria, this method showed higher sensitivity and PPV values, maintaining the same specificity and NPV. Our data show that both of the immunoassays have concordant results and can be relied on for the initial evaluation of a suspected CDAD case. According to the specific conditions within each center, the choice between these methods could be based on their specific needs. In conclusion, the most important concern about CDAD is prevention, given that many hospitals lack practice guidelines as well as diagnostic tools (19,20). In a recent study in our center, the incidence of positive fecal samples for C. difficile detection by immunoassay was 5.43% (170 out of the 3130 samples analyzed in a 5-year period), representing an incidence of CDAD of 5.04 x 1000 hospital discharges Table 2. Comparison of the results from both immunoassays and toxigenic culture in 230 fecal samples.* Number
VIDAS CDA/B
ImmunoCard A/B
Culture/tcdA/tcdB/tcdC
12 2 2 4 0 1 209
Positive Positive Positive Positive Negative Negative Negative
Positive Positive Negative Negative Positive Negative Negative
Positive Negative Positive Negative Positive Positive Negative
*Undetermined VIDAS CDA/B results considered as positives.
(21). We consider that VIDAS CDA/B might work well in large volume laboratories requiring more time from its personnel doing different tasks, and in those with a wide variety of serologic tests that could all be done with the same automated system. On the other hand ImmunoCard A/B would be more suitable in laboratories with a smaller load of samples where the personnel has more time confined to this test.
References 1. Aslam S, Musher DM. An update on diagnosis, treatment, and prevention of Clostridium difficile-associated disease. Gastroenterol Clin N Am 2006;35:315e335. 2. Loo VG, Poirier L, Miller MA, et al. A predominantly clonal multiinstitutional outbreak of Clostridium difficile-associated diarrhea with high morbidity and mortality. N Engl J Med 2005;353:2442e2449. 3. Muto CA, Pokrywka M, Shutt K, et al. A large outbreak of Clostridium difficile associated disease with an unexpected proportion of deaths and colectomies at a teaching hospital following increased fluoroquinolone use. Infect Control Hosp Epidemiol 2005;26:273e280. 4. Pe´pin J, Valiquette L, Alary ME, et al. Clostridium difficile-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ 2004;171:466e472. 5. Musher DM, Manhas A, Jain P, et al. Detection of Clostridium difficile toxin: comparison of enzyme immunoassay results with results obtained by cytotoxicity Assay. J Clin Microbiol 2007;45:2737e2739. 6. Turgeon DK, Novicki TJ, Quick J, et al. Six rapid test for direct detection of Clostridium difficile and its toxins in fecal samples compared
96
7.
8.
9.
10.
11.
12.
13.
Herrera-Ca´ceres et al./ Archives of Medical Research 41 (2010) 92e96 with the fibroblast cytotoxicity assay. J Clin Microbiol 2003;41: 667e670. Van den Berg RJ, Vaessen N, Endtz HP, et al. Evaluation of real-time PCR and conventional diagnostic methods for the detection of Clostridium difficile-associated diarrhoea in a prospective multicentre study. J Med Microbiol 2007;56:36e42. Nerandzic MM, Donskey CJ. An effective and reduced cost modified selective 1 medium for isolation of Clostridium difficile. J Clin Microbiol 2009;47:397e400. Sloan LM, Duresko BJ, Gustafson DR, et al. Comparison of real-time PCR for detection of the tcdC gene with four toxin immunoassays and culture in diagnosis of Clostridium difficile infection. J Clin Microbiol 2008;46:1996e2001. Persson S, Torpdahl M, Olsen KE. New multiplex PCR method for the detection of Clostridium difficile toxin A (tcdA) and toxin B (tcdB) and the binary toxin (cdtA/cdtB) genes applied to a Danish strain collection. Clin Microbiol Infect 2008;14:1057e1064. Clabots CR, Gerding SJ, Olson MM, et al. Detection of asymptomatic Clostridium difficile carriage by an alcohol shock procedure. J Clin Microbiol 1989;27:2386. Camacho-Ortiz A, Ponce-de-Leo´n A, Sifuentes-Osornio J. Clostridium difficile-associated disease in Latin America. Gac Med Mex 2009;145: 223e229. Fricker-Hidalgo H, Saddoux C, Suchel-Jambon AS, et al. New Vidas assay for toxoplasma-specific IgG avidity: evaluation on 603 sera. Diag Microbiol Infect Dis 2006;52:167e172.
14. Weber B, Van dar Taelem-Brule´ N, Berger A, et al. Evaluation of a new automated assay for hepatitis B surface antigen (HBsAg) detection VIDAS HBsAg Ultra. J Virol Methods 2006;135: 109e117. 15. Medici MC, Martinelli M, Albonetti V, et al. Evaluation of Rubella Virus Immunoglobulin G (IgG) and IgM Assays with the New Vidia Instrument. J Clin Microbiol 2008;46:1847e1849. 16. Iqbal J, Khalid N. Detection of acute Toxoplasma gondii infection in early pregnancy by IgG avidity and PCR analysis. J Med Microbiol 2007;56:1495e1499. 17. Miendje Deyi VY, Vandenberg O, Mascart G, et al. Diagnostic value of five commercial tests for the rapid diagnosis of Clostridium difficileassociated disease Clin Lab 2008;54:9e13. 18. Alcala´ L, Sa´nchez-Cambronero L, Catala´n MP, et al. Comparison of three commercial methods for rapid detection of Clostridium difficile toxins A and B form fecal specimens. J Clin Microbiol 2008;46: 3833e3835. 19. Macı´as AE, Ponce-de-Leo´n S. Infection control: old problems and new challenges. Arch Med Res 2005;36:637e645. 20. Planche T, Aghaizu A, Holliman R, et al. Diagnosis of Clostridium difficile infection by toxin detection kits: a systematic review. Lancet Infect Dis 2008;8:777e784. 21. Camacho-Ortı´z A, Galindo-Fraga A, Macı´as AE, et al. Factors associated with Clostridium difficile disease in a tertiary-care medical institution in Mexico: a case-control study. Rev Invest Clı´n 2009;61: 371e377.
ORIGINAL ARTICLE
Concordance Between Two Enzyme Immunoassays for the Detection of Clostridium difficile Toxins Jaime O. Herrera-Ca´ceres, Adria´n Camacho-Ortiz, Arturo Galindo-Fraga, Melissa Herna´ndez-Dura´n, Andrea Cordero-Rangel, Araceli Herna´ndez-Cruz, Miriam Bobadilla-del Valle, Alfredo Ponce-de-Leo´n, and Jose´ Sifuentes-Osornio Laboratory of Clinical Microbiology, National Institute of Medical Sciences and Nutrition Salvador Zubira´n, Mexico City, Mexico Received for publication October 21, 2009; accepted January 26, 2010 (ARCMED-D-09-00509).
Background and Aims. The diagnosis of Clostridium difficile-associated disease (CDAD) is based on the detection of toxins from stool samples. There are several immunoassays for this purpose. The aim of this study was to determine the concordance between the two immunoassays and their performance in comparison to the toxigenic culture as part of the initial evaluation of a suspected case of CDAD. Methods. All fecal samples submitted for detection of C. difficile toxins during a 5-month period to our laboratory were analyzed by two immunoassays, VIDAS Toxin CDA/B assay (BioMerieux) and ImmunoCard Toxins A/B (Meridian Bioscience). We cultured on cycloserine-cefoxitin-fructose agar and PCR was used for detection of toxigenic genes. Real-time PCR was performed directly from samples to detect the tcdC gene. Results. At the end of the study we processed 230 samples, 13 were positive using VIDAS CDA/B (5.6%), and 14 using ImmunoCard A/B (6.0%); k coefficient was 0.857. With ImmunoCard A/B we obtained a sensitivity of 80%, a specificity of 99%, positive predicitive value (PPV) 86% and negative predictive value (NPV) 98%, as compared to toxigenic culture. For VIDAS CDA/B we obtained a sensitivity of 90%, a specificity of 98%, PPV 69% and NPV 99%, compared to the same standard. There were seven undetermined results (3.0%) by VIDAS CDA/B. Five of these had a positive culture and all the patients had symptoms of CDAD. Considering these undetermined results as positive, we calculated a sensitivity of 93%, specificity of 97%, PPV 71% and NPV of 99% for this test, and a k of 0.856. Both immunoassays showed similar results and are suitable for the initial evaluation of patients with suspected CDAD. Conclusions. Our data suggest that an undetermined result of VIDAS CDA/B should be considered as positive if CDAD is suspected. Additionally, both immunoassays showed similar results and are suitable for the initial evaluation of patients with suspected CDAD. Ó 2010 IMSS. Published by Elsevier Inc. Key Words: Diarrhea, Toxins, Clostridium difficile, Nosocomial infections.
Introduction Approximately 3e29% of patients receiving antibiotics will develop diarrhea; moreover, Clostridium difficile will be the culprit agent in up to 10e25% of these cases (1). Address reprint requests to: Jose´ Sifuentes-Osornio, MD, Laboratorio de Microbiologı´a Clı´nica, Instituto Nacional de Ciencias Me´dicas y Nutricio´n Salvador Zubira´n, Vasco de Quiroga #15, C.P. 14000, Tlalpan, Me´xico D.F. Me´xico; Phone: (þ52) (55) 5487-0900, Ext. 2174; FAX: (þ52) (55) 5513-3945; E-mail: [email protected]
In recent years, complications and mortality associated with this bacteria have increased dramatically (2e4). To date, the cytotoxicity assay for detection of toxin B is considered the gold standard for diagnosis of C. difficile-associated disease (CDAD). However, this procedure is cumbersome, costly and the results are not available before 48 h. Several immunoassays have been developed in order to evaluate the presence of both A and B toxins in stool samples as a simpler and faster option (5). Two popular immunoassays are the VIDAS Clostridium difficile Toxin A/B assay
0188-4409/$esee front matter. Copyright Ó 2010 IMSS. Published by Elsevier Inc. doi: 10.1016/j.arcmed.2010.03.004
Immunoenzymatic Detection of Clostridium difficile toxins
(bioMerieux, Marcy l’Etoile, France) with good sensitivity (88.3%) and specificity (99.8%) values, a positive predictive value (PPV) of 98.1%, and a negative predictive value (NPV) of 98.4%, and the ImmunoCard Toxins A/B (Meridian Bioscience, Cincinnati, OH), which have shown a sensitivity of 55.4%, a specificity of 92.2%, a PPV of 44.4%, and a NPV of 94.9%, when compared with tissue culture cytotoxin B assay (CBA) (6,7). Stool culture for C. difficile is time consuming and requires anaerobic conditions as well as a special medium containing cycloserincefoxitin-fructose agar (CCFA) or the recently studied C. difficile-Brucella agar (8,9). The presence of C. difficile can be determined by other methods including multiplex polymerase chain reaction (M-PCR) to detect tcdA and tcdB genes from culture (toxigenic culture) (10) and realtime PCR (RT-PCR) directly from stool samples to detect the tcdC gene (regulatory gene) with good sensitivity and specificity (9). The aim of this study was to determine the concordance between the two immunoassays and their performance in comparison to the toxigenic culture as part of the initial evaluation of a suspected case of CDAD. Materials and Methods Setting and Time Period During a 5-month period (MaySeptember, 2008), all samples sent for detection of C. difficile toxins to the Laboratory of Clinical Microbiology of a 200-bed tertiary-care medical center located in Mexico City were analyzed by both immunoassays VIDAS CDA/B (bioMerieux) and ImmunoCard A/B (Meridian Bioscience) methods. Each immunoassay was performed following manufacturer recommendations. Results from VIDAS CDA/B were reported in relative fluorescence values (RFV). Culture Methods All stool samples were cultured for C. difficile using CCFA and blood sheep agar (BSA) medium in anaerobic conditions after alcohol shock (11). Upon arrival, we stored 2 g of feces at 80 C. The positive cultures were identified by Vitek ANI test cards (bioMerieux). Molecular Methods The isolates were tested for the presence of tcdA and tcdB genes using a modification of a PCR-method previously described (9). In short, the sequences of the primers used for amplification were tcdB-F, GAGCTGCTTCAATTGGAGAGA; tcdB-R, GTAACCTACTTTCATAACACCAG; tcdA-F, ATGATAAGGCAACTTCAGTGG, and tcdA-R, TAAGTTCCTCCTGCTCCATCAA. The cycling parameters were amplification of fragments of the tcdA gene (624 bp) and the tcdB gene (412 bp) by conventional PCR; the conditions of the multiplex PCR reaction were
93
final volume 25 ml, dNTPs 300 mM, MgCl2 3 mM, primers 25 pM and Taq polymerase 2.5 units; the cycling parameters were denaturation for 5 min at 94 C and amplification consisting of 1 min at 94 C, 1 min at 50 C, and 1 min at 72 C for 30 cycles. The presence of an amplified product was confirmed by agarose gel detection. Real-time PCR was used to confirm the presence of the tcdC gene from the frozen stool as previously described (9). This method has previously shown high sensitivity (86%) and specificity (97%). Briefly, this PCR assay detects the presence of the tcdC gene as well as the 18-bp, 39-bp, and other deletions found in the tcdC gene of C. difficile. We used the kit 296 (TIB MolBiol LLC, Adelphia, NJ) following the manufacturer’s recommendations. Positive and negative controls were included with each run. The positive control was C. difficile ATCC 9689 (wild type) and a clinical isolate C. difficile type 027 kindly provided by Dr. Ellen J. Baron with the 18-bp deletion. The negative control was a recent clinical isolate of C. sordelli. Ethics and Statistical Analysis This study was evaluated and approved by the Institutional Review Board. For concordance analysis we excluded samples with an insufficient amount of stool for proper analysis by both methods. We also excluded repeated samples from the same patient. Sensitivity, specificity, positive and negative predictive values were calculated. Coefficient of k was considered useful with a value $0.4. We used SPSS software v. 15. Results A total of 234 fecal samples were processed during the study period. Of these, four were discarded: two due to an insufficient amount of feces and two for being repeated samples. Two hundred and thirty samples were included in the final analysis. Using the ImmunoCard A/B 14 (6%) samples were determined as positive and with the VIDAS CDA/B 13 samples were positive (5.6%). Of the 230 cultures included in the study, only 13 had bacterial growth (5.6%). The presence of tcdA and/or tcdB genes were confirmed in the 13 isolates by M-PCR (one isolate was tcdA negative but tcdB positive). Real-time PCR showed the presence of the tcdC gene directly from 15 fecal samples, two of them with a negative culture; therefore M-PCR could not be done. With these results, we compared the ImmunoCard A/B and VIDAS CDA/B with the results from the anaerobic culture and the presence of tcdC gene. Using the latter as the standard method, we calculated sensitivity, specificity, PPV and NPV. The ImmunoCard A/B test showed 80, 99, 86, and 99%, respectively, and the VIDAS CDA/B test showed 90, 98, 69 and 99%, respectively. VIDAS CDA/B values were classified into three categories: positive, negative and undetermined as recommended by the manufacturer. Two hundred and ten samples
94
Herrera-Ca´ceres et al./ Archives of Medical Research 41 (2010) 92e96
were negative, seven were undetermined and 13 were positive. Then, the seven samples with undetermined results were repeated. Of these seven samples, two were undetermined by a second assay, one was positive, and four were negative upon retesting. A more detailed analysis of these seven patients found that all had diarrhea days before the diagnostic test (median 2 days, range 2e15 days) was processed, four had fever, one had an elevated peripheral white blood cell count, and all had a normal serum creatinine. All had also received antibiotics for at least 3 days during the 6 weeks prior to sample processing and were hospitalized for a minimum of 2 days, two were hospitalized at the intensive care unit, and five were receiving proton pump inhibitors. Additionally, we compared these seven samples with the results from the other diagnostic tools used, finding a positive culture in five samples. The presence of the toxinproducing genes was confirmed by the M-PCR assay for these five cultures (Table 1). Distribution of the initial RFV of the three groups according to the VIDAS CDA/B method is shown in Figure 1. The group of samples with a negative value showed a mean value and the one standard deviation values far below the negative cutoff value. Samples with an undetermined result had a trend towards higher values close to the positive cutoff value, and samples with a positive result showed values far above the positive cutoff value. With the initial results the k coefficient between ImmunoCard A/B test and VIDAS CDA/B test was 0.857, and after the inclusion of the seven undetermined VIDAS CDA/B results into the positive group, it was 0.856. When considering VIDAS CDA/B undetermined results as positive, we obtained a sensitivity of 93%, a specificity of 97%, PPV of 71% and NPV of 99%. In Table 2 we describe the performance of the three methods used in this trial. The numbers of all the possible combined results by VIDAS CDA/B, ImmunoCard A/B and toxigenic culture results are shown. We found 12 samples that were positive by all methods, four others had a positive VIDAS CDA/B result but both of the other methods were negative. Two samples had a positive result by both immunoassays but neither culture nor RT-PCR confirmed these results. Two more samples were positive by VIDAS CDA/B and by toxigenic culture but negative using ImmunoCard A/B. From all samples there was only one positive toxigenic culture that had a negative result by both immunoassays. Discussion CDAD has gained great importance during the present decade. Detection of toxins in stool samples is fundamental for the diagnosis (1). However, a significant number of hospitals in Mexico and in many other developing countries still lack a laboratory method for such a purpose. As previously reported, less than one third of the Latin American
Table 1. Description of the results of those 7 samples with an undetermined value using the VIDAS CDA/B test
Vidas CDA/B 1. Undetermined*/ Undetermined** 2. Undetermined*/ Undetermined** 3. Undetermined*/Positive** 4. Undetermined*/Negative** 5. Undetermined*/Negative** 6. Undetermined*/Negative** 7. Undetermined*/Negative**
Immunocard A/B
Toxigenic Culture
Negative
Positive
Positive
Positive
Positive Negative Positive Negative Negative
Positive Positive Positive Negative Negative
*First assay. **Re-test.
countries (7/25 countries recognized by the Pan-American Health Organization) have reported cases of CDAD (12). The results of this study show the performance and a high concordance (kappa 0.856) among two easy to perform immunoassays. Although we observed a moderate prevalence of CDAD (6.52%) during this study, both of these immunoassays were quite trustworthy when compared with a good standard using up-to-date molecular biotechnology (toxigenic culture). VIDAS CDA/B, has the advantage of quantitative results of fluorescence that are automatically interpreted as positive, negative or undetermined. Nevertheless, in order to use VIDAS CDA/B it is necessary to have the VIDAS automated system, that can be used for many other serologic tests and has proven to be trustful for these other tests (13e16). On the other hand ImmunoCard A/B is a methodologically simple that does not need any special equipment other than a standard 4e8 C refrigerator and a vortex mixer. For these reasons and the prompt results (up to 30 min), ImmunoCard A/B can be used by non-specialized personnel and needs no sophisticated laboratory environment. It has been also widely studied and has a proven credibility detecting both toxins, A and B (17,18). It has also been shown that well-type assays and membrane-type assays to detect these toxins do not differ significantly when compared with cytotoxicity as a reference standard. Welltype assays and membrane-type assays with glutamate dehydrogenase as a target perform equally well and have good correlation with feces culture of C. difficile. Furthermore in our study, the results by this method were corroborated by two observers. In three samples there was discordance among the observers and the test was repeated and given as negative by both on the second trial. Given the results from the sub analysis of the RFV obtained by the VIDAS CDA/B, it would be advisable to treat such a patient with specific therapy anti-CDAD if this patient has risk factors, develops diarrhea and has an undetermined result by the VIDAS CDA/B test. Using these
Relative Fluorescence Values
Immunoenzymatic Detection of Clostridium difficile toxins
95
6.22 6.07 6.01 4.15 3.94 1.95 1.53 1.41 1.39 1.37 0.53 0.51 0.49 0.34 0.32 0.24 0.21 0.2 0.16 0.13 0.11 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01
Negative
Undetermined
Positive
VIDAS CDA/B Result Figure 1. Distribution of relative florescence values from the VIDAS CDA/B Cutoff relative fluorescence values: Negative #14 units, Undetermined O14e37 units, Positive O37 units, as shown by the red doted lines. The mean value of the negative results was 0.02 (minimum 0.01- maximum 0.13, SD 0.022), the mean value of the undetermined results was 0.24 (0.16e0.34, SD 0.06), and the mean value of the positive results was 1.41 (0.49e6.22, SD 2.23).
criteria, this method showed higher sensitivity and PPV values, maintaining the same specificity and NPV. Our data show that both of the immunoassays have concordant results and can be relied on for the initial evaluation of a suspected CDAD case. According to the specific conditions within each center, the choice between these methods could be based on their specific needs. In conclusion, the most important concern about CDAD is prevention, given that many hospitals lack practice guidelines as well as diagnostic tools (19,20). In a recent study in our center, the incidence of positive fecal samples for C. difficile detection by immunoassay was 5.43% (170 out of the 3130 samples analyzed in a 5-year period), representing an incidence of CDAD of 5.04 x 1000 hospital discharges Table 2. Comparison of the results from both immunoassays and toxigenic culture in 230 fecal samples.* Number
VIDAS CDA/B
ImmunoCard A/B
Culture/tcdA/tcdB/tcdC
12 2 2 4 0 1 209
Positive Positive Positive Positive Negative Negative Negative
Positive Positive Negative Negative Positive Negative Negative
Positive Negative Positive Negative Positive Positive Negative
*Undetermined VIDAS CDA/B results considered as positives.
(21). We consider that VIDAS CDA/B might work well in large volume laboratories requiring more time from its personnel doing different tasks, and in those with a wide variety of serologic tests that could all be done with the same automated system. On the other hand ImmunoCard A/B would be more suitable in laboratories with a smaller load of samples where the personnel has more time confined to this test.
References 1. Aslam S, Musher DM. An update on diagnosis, treatment, and prevention of Clostridium difficile-associated disease. Gastroenterol Clin N Am 2006;35:315e335. 2. Loo VG, Poirier L, Miller MA, et al. A predominantly clonal multiinstitutional outbreak of Clostridium difficile-associated diarrhea with high morbidity and mortality. N Engl J Med 2005;353:2442e2449. 3. Muto CA, Pokrywka M, Shutt K, et al. A large outbreak of Clostridium difficile associated disease with an unexpected proportion of deaths and colectomies at a teaching hospital following increased fluoroquinolone use. Infect Control Hosp Epidemiol 2005;26:273e280. 4. Pe´pin J, Valiquette L, Alary ME, et al. Clostridium difficile-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ 2004;171:466e472. 5. Musher DM, Manhas A, Jain P, et al. Detection of Clostridium difficile toxin: comparison of enzyme immunoassay results with results obtained by cytotoxicity Assay. J Clin Microbiol 2007;45:2737e2739. 6. Turgeon DK, Novicki TJ, Quick J, et al. Six rapid test for direct detection of Clostridium difficile and its toxins in fecal samples compared
96
7.
8.
9.
10.
11.
12.
13.
Herrera-Ca´ceres et al./ Archives of Medical Research 41 (2010) 92e96 with the fibroblast cytotoxicity assay. J Clin Microbiol 2003;41: 667e670. Van den Berg RJ, Vaessen N, Endtz HP, et al. Evaluation of real-time PCR and conventional diagnostic methods for the detection of Clostridium difficile-associated diarrhoea in a prospective multicentre study. J Med Microbiol 2007;56:36e42. Nerandzic MM, Donskey CJ. An effective and reduced cost modified selective 1 medium for isolation of Clostridium difficile. J Clin Microbiol 2009;47:397e400. Sloan LM, Duresko BJ, Gustafson DR, et al. Comparison of real-time PCR for detection of the tcdC gene with four toxin immunoassays and culture in diagnosis of Clostridium difficile infection. J Clin Microbiol 2008;46:1996e2001. Persson S, Torpdahl M, Olsen KE. New multiplex PCR method for the detection of Clostridium difficile toxin A (tcdA) and toxin B (tcdB) and the binary toxin (cdtA/cdtB) genes applied to a Danish strain collection. Clin Microbiol Infect 2008;14:1057e1064. Clabots CR, Gerding SJ, Olson MM, et al. Detection of asymptomatic Clostridium difficile carriage by an alcohol shock procedure. J Clin Microbiol 1989;27:2386. Camacho-Ortiz A, Ponce-de-Leo´n A, Sifuentes-Osornio J. Clostridium difficile-associated disease in Latin America. Gac Med Mex 2009;145: 223e229. Fricker-Hidalgo H, Saddoux C, Suchel-Jambon AS, et al. New Vidas assay for toxoplasma-specific IgG avidity: evaluation on 603 sera. Diag Microbiol Infect Dis 2006;52:167e172.
14. Weber B, Van dar Taelem-Brule´ N, Berger A, et al. Evaluation of a new automated assay for hepatitis B surface antigen (HBsAg) detection VIDAS HBsAg Ultra. J Virol Methods 2006;135: 109e117. 15. Medici MC, Martinelli M, Albonetti V, et al. Evaluation of Rubella Virus Immunoglobulin G (IgG) and IgM Assays with the New Vidia Instrument. J Clin Microbiol 2008;46:1847e1849. 16. Iqbal J, Khalid N. Detection of acute Toxoplasma gondii infection in early pregnancy by IgG avidity and PCR analysis. J Med Microbiol 2007;56:1495e1499. 17. Miendje Deyi VY, Vandenberg O, Mascart G, et al. Diagnostic value of five commercial tests for the rapid diagnosis of Clostridium difficileassociated disease Clin Lab 2008;54:9e13. 18. Alcala´ L, Sa´nchez-Cambronero L, Catala´n MP, et al. Comparison of three commercial methods for rapid detection of Clostridium difficile toxins A and B form fecal specimens. J Clin Microbiol 2008;46: 3833e3835. 19. Macı´as AE, Ponce-de-Leo´n S. Infection control: old problems and new challenges. Arch Med Res 2005;36:637e645. 20. Planche T, Aghaizu A, Holliman R, et al. Diagnosis of Clostridium difficile infection by toxin detection kits: a systematic review. Lancet Infect Dis 2008;8:777e784. 21. Camacho-Ortı´z A, Galindo-Fraga A, Macı´as AE, et al. Factors associated with Clostridium difficile disease in a tertiary-care medical institution in Mexico: a case-control study. Rev Invest Clı´n 2009;61: 371e377.